The Borg Lab

GTSAM 2.3.1 Changes (from 2.3.0)

GTSAM 2.3.1 is a minor update from 2.3.0.  This release makes improvements to the IMU summary factors.  These factors allow the use of high-rate IMUs in smoothing by summarizing many IMU measurements into one, in a way that still permits efficient and accurate relinearization and estimation of biases, closely following the methods of Lupton and Sukkarieh in TRO 2012.  Specific changes include:

  • IMU summary factors now use analytic derivatives for further speedup.
  • Simplified IMU factor API - see ImuFactor andCombinedImuFactor, in the gtsam/navigation subdirectory.
  • Example of using the new IMU factor with the KITTI dataset, using the GTSAM MATLAB toolbox, IMUKittiExampleGPS.m.
  • Fixed indexing bug in KalmanFilter that made solving slow down over time.
  • Upgraded internal Eigen version to 3.2.0.
  • Other small API and build system improvements.

Older change lists: